How the SolWeather System Works

A plain-language guide for members

SolWeather isn’t just a set of weather maps — it’s a structured, multi-layered forecasting system designed to help Australian farmers make informed, forward-looking seasonal decisions. Our approach blends long-range climate cycles, AI-assisted interpretation, and real-time model data into a single forecasting framework.

1. Long-Range Forecasting: Cycles, AI and Climate History

At the core of SolWeather is a proprietary cyclical algorithm developed in-house. It analyses more than 124 years of historical Australian rainfall data to identify repeatable climate patterns and long-wave seasonal signals that influence winter and spring outcomes.

Our long-range probability forecasts (monthly, 3-month, 6-month, and 12-month) are generated from mean rainfall outcomes during past analogue cycles. However, like any historical dataset, this includes outlier years and anomalies that don’t always align with the dominant pattern.

To improve this, we are actively developing a data control system to better filter non-relevant anomalies and refine the reliability of our analogue signal base. This project is ongoing and will continue to enhance our long-term forecast accuracy.

To support and interpret the maps, we use AI tools to scan the long-range outputs each month. The AI assists in identifying regional risks, rainfall anomalies, and potential timing shifts — generating first-pass commentary that is then reviewed by the SolWeather team.

2. The SolWeather Long-Range Map Suite

Each month, we publish a complete suite of 17 long-range rainfall probability maps:

  • 12 monthly outlooks (April through March)

  • 4 seasonal outlooks (e.g. Apr–Jun, May–Jul, Jun–Aug, Jul–Sep)

  • 2 six-month outlooks (e.g. Apr-Sep, Oct-Mar)

  • 1 twelve-month outlook (e.g. April 2025 to March 2026)

Each map shows the likelihood of exceeding median rainfall across Australia’s cropping zones and is accompanied by regional commentary informed by both AI and human interpretation.

3. Layered Forecasting for Confidence and Context

SolWeather does not rely on a single model or timeframe. Instead, we use a layered approach to forecasting by combining:

  • 5-day short-term models

  • 16-day model outlooks and frontal progression

  • 35–46 day extended model guidance

  • Soil moisture data and intra-regional trend detection

Each layer is evaluated against the long-range baseline to determine how short-term and medium-range changes fit into the broader seasonal pattern.

4. Region-Specific Daily Forecasts

Each day, SolWeather members receive updated regional forecasts tailored to major agricultural zones across WA, SA, VIC, NSW and QLD. These include:

  • Short-term rainfall, temperature, wind and gust summaries

  • Current soil moisture conditions and rainfall timing outlooks

  • AI-generated regional forecast commentary

  • Guidance for sowing, spraying, grazing, and managing seasonal risk

  • Confidence levels based on model alignment and long-range signal support

Why This System Works

We don’t claim perfection — but we do aim for clarity, consistency, and actionable insight. SolWeather’s system is built to help reduce uncertainty, improve timing, and give members a clearer view of the season ahead.

We update long-range maps monthly, interpret forecasts daily, and continue refining how we deliver rainfall probabilities, confidence levels, and planning notes that support informed agricultural decision-making.